Key points are not available for this paper at this time.
Two important problems encountered in X-ray astronomy are considered: the assessment of the validity of models, and the determination of values of adjustable parameters which valid models possess. Minimization of the chi-squared statistic is recommended for testing the validity of classes of models. The chi-squared min + 1 error estimation method is evaluated and found to be unsuitable when several parameter ranges are to be derived. The ratio of variances method gives parameter confidence regions which are more variable than necessary.
Lampton et al. (Sun,) studied this question.